ax = msno.matrix(app.df)
%pylab inline
import pandas as pd
import missingno as msno
import seaborn as sns
Populating the interactive namespace from numpy and matplotlib
%load_ext autoreload
%autoreload 2
# configurations pandas
pd.set_option('display.max_rows', 250)
pd.set_option('display.max_columns', 250)
from src.dataapp import App
app = App()
app.load()
_ = app.aggregate_by_id()
Loading ./data/tables/application_train.csv ./data/tables/application_train.csv loaded (307511, 122) Loading ./data/tables/application_test.csv ./data/tables/application_test.csv loaded (48744, 121) Loading ./data/tables/bureau_balance.csv ./data/tables/bureau_balance.csv loaded (27299925, 3) Loading ./data/tables/bureau.csv ./data/tables/bureau.csv loaded (1716428, 17) Loading ./data/tables/credit_card_balance.csv ./data/tables/credit_card_balance.csv loaded (3840312, 23) Loading ./data/tables/installments_payments.csv ./data/tables/installments_payments.csv loaded (13605401, 8) Loading ./data/tables/POS_CASH_balance.csv ./data/tables/POS_CASH_balance.csv loaded (10001358, 8) Loading ./data/tables/previous_application.csv ./data/tables/previous_application.csv loaded (1670214, 37) Aggregate by ID application Shape après aggregation : (356255, 247) Aggregate by ID bureau Aggregate by ID bureau_balance Shape après aggregation : (356255, 359) Aggregate by ID previous_application Shape après aggregation : (356255, 592) Aggregate by ID pos_cash_balance Shape après aggregation : (356255, 609) Aggregate by ID installments_payments Shape après aggregation : (356255, 635) Aggregate by ID credit_card_balance Shape après aggregation : (356255, 771) Shape finale : (356255, 771)
ax = msno.matrix(app.df)
print('Positive: {}'.format(app.df[app.df['TARGET'] == 1].shape))
print('Negative: {}'.format(app.df[app.df['TARGET'] == 0].shape))
Positive: (24825, 771) Negative: (282686, 771)
plt.figure(figsize=(15,10))
ax = sns.heatmap(app.df.sample(frac=0.1).corr(method='pearson'), cmap='RdYlGn_r', xticklabels=False, yticklabels=False)
plt.figure(figsize=(15,10))
ax = sns.histplot(data=app.df, x="PAYMENT_RATE", hue="TARGET", bins=50)
plt.figure(figsize=(15,10))
ax = sns.histplot(data=app.df, x="EXT_SOURCE_1", hue="TARGET", bins=50)
plt.figure(figsize=(15,10))
ax = sns.histplot(data=app.df, x="EXT_SOURCE_2", hue="TARGET", bins=50)
plt.figure(figsize=(15,10))
ax = sns.histplot(data=app.df, x="EXT_SOURCE_3", hue="TARGET", bins=50)
columns = ['TARGET', 'PAYMENT_RATE', 'EXT_SOURCE_1', 'EXT_SOURCE_2', 'EXT_SOURCE_3']
sns.pairplot(app.df[columns].sample(frac=0.1), hue='TARGET', diag_kind='hist', palette='tab10', height=5)
<seaborn.axisgrid.PairGrid at 0x1a7dff9d730>
filename = 'Home_Credit_Default_Risk.csv'
app.clean()
app.cleaned_df.shape
app.save(filename)
Clean data... Nombre de colonnes: 771 Le pourcentage moyen de valeurs manquantes est de 25.55% Nombre de colonnes avec un taux de remplissage minimum de 90.00%: 424 Drop empty columns Replace infinite value by NAN Save data...
msno.matrix(app.cleaned_df)
<AxesSubplot:>
!jupyter nbconvert --config nbconvert/config_html.py
[NbConvertApp] Converting notebook P7_01_notebookexploration.ipynb to html
[NbConvertApp] Writing 4623390 bytes to P7_01_notebookexploration.html
[NbConvertApp] Converting notebook P7_02_notebookscoring.ipynb to html
[NbConvertApp] Writing 4455319 bytes to P7_02_notebookscoring.html
[NbConvertApp] Converting notebook P7_03_notebookdashboard.ipynb to html
C:\Users\flori\miniconda3\envs\jupyter\lib\site-packages\nbconvert\filters\datatypefilter.py:39: UserWarning: Your element with mimetype(s) dict_keys(['application/vnd.plotly.v1+json']) is not able to be represented.
warn("Your element with mimetype(s) {mimetypes}"
[NbConvertApp] Writing 901023 bytes to P7_03_notebookdashboard.html